Mr. James Campbell reports

ROCKWELL ISSUES UPDATE ON RESOURCE ESTIMATES AND OTHER DEVELOPMENTS IN THE MIDDLE ORANGE RIVER DIAMOND FIELDS

Rockwell Diamonds Inc. has provided an update on resources relating to its proposed acquisition of the Bondeo properties along with other developments relating to its Middle Orange River strategy.

Highlights:

• 7.1-million-cubic-metre inferred mineral resource estimated at Remhoogte at a grade of 0.9 carat per hundred cubic metres and average carat value of $2,900 (U.S.) per carat;
• Consolidation of Middle Orange River expansion opportunities boosted by exploration agreement covering the Lanyonvale 376 property;
• Continuing exploration in the region initiated for fiscal 2016 to fiscal 2017;
• A high-value 121-carat yellow octagonal diamond recovered at Saxendrift.

Remhoogte resource

An inferred mineral resource of 7,056,000 cubic metres of Rooikoppie gravel at a grade of 0.9 carat per hundred cubic metres with an average value of $2,900 (U.S.) per carat has been estimated at Remhoogte after re-evaluating historical, geological and field mapping data; the survey of areas mined since commencement of operations on the property and the available production and sales results. The largest stone recovered from Remhoogte to date weighed 178 carats. The company has filed on SEDAR a technical report titled "Technical Report on the Remhoogte/Holsloot Alluvial Diamond Project, Hay District, Republic of South Africa for Rockwell Diamonds Inc.," dated March 9, 2015, with an effective date of Feb. 6, 2015, pursuant to National Instrument 43-101 in connection with the inferred mineral resource estimate for the Remhoogte resource.

In addition, as part of its continuing evaluation of the Bondeo acquisition properties, the company has identified an exploration target with potential for a further three million to four million cubic metres of fluvial-alluvial gravels (target grades of approximately 0.4 to 1.5 carats per hundred cubic metres and anticipated diamond values ranging from $2,500 (U.S.) to $4,000 (U.S.) per carat). Limited sampling of the fluvial-alluvial gravel has recovered 446 carats with the largest diamond recovered being a 27-carat stone. The company intends to announce a further resource estimation on the fluvial-alluvial gravel once bulk sampling has been completed.

Exploration agreement at Lanyonvale 376

As part of its continuing exploration of the MOR diamond potential, the company announces that it has entered into an exploration agreement which covers the Lanyonvale property which is considered to be transgressed by paleo river channels with significant potential to add to the company's resources and mining flexibility.

An exploration agreement reached with the prospecting and surface rights holder of the 2,300-hectare property provides for the company to determine the potential of the property which is contiguous to the Western boundary of Wouterspan and the Northern boundary of Niewejaarskraal and subject to positive results will include the property in its mining plans.

Exploration update

The company is continuing its exploration program in the region and announces that in fiscal 2016, it plans to spend $350-million (four million rand) on an exploration program consisting of field mapping and geological delineation, Rooikoppie pitting and fluvial-alluvial drilling across its properties, at Niewejaarskraal and Wouterspan. At Lanyonvale, a combined program of mapping, pitting and reverse circulation drilling is planned which could potentially lead to the announcement of an exploration target. Based on the positive results of the preliminary economic study at Wouterspan, pitting will be carried out on a tight grid to add to the geological understanding upon the recommencement of trial mining.

121-carat rough diamond recovered at Saxendrift

Rockwell is pleased to announce that a high-value 121-carat yellow octagonal rough diamond was recovered at Saxendrift, the company's flagship operation in the MOR in February, 2015, bringing the number of plus-100-carat diamonds recovered in the MOR to 10 since September, 2013.

Commenting on the transaction James Campbell, chief executive officer and president, said: "Estimation of an inferred resource at Remhoogte has the potential to add approximately 63,500 carats to our diamond inventory, at an average value of $2,900 (U.S.) per carat. This further entrenches Rockwell's positioning in the MOR alluvial diamond fields. The Lanyonvale exploration agreement, together with the recently granted mining and prospecting rights in the MOR, will provide additional greenfields opportunities to support our longer-term organic growth objectives. In the short term, our intended acquisition of the Remhoogte/Holsloot project is expected to lift our monthly volume beyond our 500,000-cubic-metre target. With the additional MOR exploration and development projects, there is potential for us to be able to further grow our production base and improve the company's profitability post the Bondeo acquisition. We have launched a formal exploration program, underpinned by our in-depth knowledge of the regional geology, to ensure that we deliver these new opportunities as projects."

Additional information on the Remhoogte/Holsloot project

Sampling

Steyn Diamante CC has been operating on the Remhoogte/Holsloot project since March, 2014. Steyn has sampled mainly colluvial Rooikoppie gravels from the properties. The diamond register indicates that about 10,379.48 carats were recovered from the Remhoogte/Holsloot project during the period from March 17, 2014, to Feb. 6, 2015. These stones were sold on the open market for about $2,900 (U.S.) per carat.

Due to the nature of the Steyn operation on the Remhoogte/Holsloot project, only limited verification of the sample data has been possible.

All diamonds recovered have been recorded by Steyn and a list was provided to Rockwell in an Excel spreadsheet. It has not been possible to verify the accuracy of these data. It has, further, not been possible to determine the efficiencies of the processing plants as details of the tracer tests are not available (although such tests were done, the results were never recorded).

Payment for diamond parcels was received by electronic transfer and a formal broker's note was provided from the buyer. These data are, subsequently, added to the production database. Each broker's note has been examined by the qualified persons responsible for the technical report on the Remhoogte/Holsloot project and found to be present and correct.

The drill hole database was imported into Rockwell's Datashed QAQC, a database program which enforces compliance through various protocol levels. This process along with an independent review of the database highlighted various inaccuracies in 45 holes. A reassessment of these holes by Rockwell indicated that the errors were present in the original data. As a result, these holes have been excluded from any further interpretation. With respect to the drilling data, Rockwell's previous encounters have shown that there is often less than 80-per-cent correlation between the drilling results and actual (fluvial-alluvial) gravel profiles as seen in the mining/sampling pits. It is important to note that this lack of confidence in the drilling results extends primarily to the calcreted fluvial-alluvial units. The identification of the Rooikoppie gravel units is a much simpler matter of bright red sandy gravels versus white calcreted gravels.

The surveyed sample volumes were sent to Rockwell under certificate from an independent professional surveyor (RBG Surveys Pty. Ltd.) and subsequently captured onto the production database by the database co-ordinator and verified by the group technical manager. The procedures and protocols of the surveyor have been reviewed by the independent qualified person and found to be in accord with industry standard. In addition, the surveyor provided a digital terrain model of the area surveyed to Rockwell. This DTM was imported into Gemcom Surpac (v 6.0.1), which calculated the volume of the mined area -- no material difference was found between the calculated volume and the survey volume.

During January, 2015, Rockwell appointed RBG to survey the areas sampled by Steyn. The surveyed sample volumes as at Feb. 6, 2015, totalled 1,123,877 cubic metres.

Mineral resource estimates

The Canadian Institute of Mining Metallurgy and Petroleum definitions do not deal specifically with the peculiarity of alluvial diamonds deposits when it comes to mineral resource or reserve estimations. The reason for this is that, historically, companies mining such deposits have not been listed on any Canadian exchanges requiring guidance in these matters. The South African Code for Reporting of Mineral Resources and Mineral Reserves, however, has separate definitions for diamond resources and reserves. The inferred mineral resource categories used in the news release follows the CIM definition. The resultant estimations are materially similar to those set out in the SAMREC Code CIM Standards define inferred mineral resources as "... that part of a Mineral Resource for which quantity and grade or quality can be estimated on the basis of geological evidence and limited sampling and reasonably assumed, but not verified, geological and grade continuity."

The estimate is based on limited information and sampling gathered through appropriate techniques from locations such as outcrops, trenches, pits, workings and drill holes. Due to the uncertainty that may be attached to inferred mineral resources, it cannot be assumed that all or any part of an inferred mineral resource will be upgraded to an indicated or measured mineral resource as a result of continued exploration. Confidence in the estimate is insufficient to allow the meaningful application of technical and economic parameters or to enable an evaluation of economic viability worthy of public disclosure. Inferred mineral resources must be excluded from estimates forming the basis of feasibility or other economic studies.

For comparison, the SAMREC code defines an inferred (diamond) resource as: "... that part of a Diamond Resource for which tonnage or volume, grade and average diamond value can be estimated with a low level of confidence. It is inferred from geological evidence and assumed, but not verified, geological and grade continuity and a sufficiently large diamond parcel is not available to ensure a reasonable representation of the diamond assortment. It is based on information gathered through appropriate techniques from locations such as outcrops, trenches, pits, workings and drill holes that may be limited or of uncertain quality and reliability."

This category, which has a lower level of confidence than that applying to an indicated mineral resource is intended to cover situations where a mineral concentration or occurrence has been identified and limited measurements and sampling completed, but where the data are insufficient to allow the geological and/or grade continuity to be confidently interpreted. Due to the uncertainty which may be attached to some inferred mineral resources, it cannot be assumed that all or part of an inferred mineral resource will necessarily be upgraded to an indicated or measured mineral resource as a result of continued exploration. Further, confidence in the estimate is insufficient to allow the meaningful application of technical and economic parameters or to enable an evaluation of economic viability worthy of public disclosure.

Current mineral resource estimates

Contrary to popular opinion, alluvial diamond deposits can effectively be evaluated according to international classifications -- they simply require an understanding of the geological and economic parameters that are peculiar to these deposits. Rockwell has applied the following criteria for the classification of inferred mineral resource estimates:

• A minimum drill grid on 100 by 100 metres;
• A minimum of 500 carats from the property (for grade and value estimation);
• An inferred resource may extend beyond the last drill holes to a distance of one-half of the nominal drill-grid spacing (that is, 50 metres), given that there is reasonable grounds to assume geological continuity;
• For grade to be considered as inferred, a minimum of 0.5 per cent to 1 per cent of the inferred volume of gravel has to be processed through a prospecting or production plant.

Rockwell's standard operating procedure to estimate a diamond resource -- volume, grade and average value include the following.

Volume

The applicable resource volume was calculated (from the original borehole data) in Gemcom Surpac (v 6.0.1) -- Rockwell has not relied on any previous estimation of volume. All of the borehole data were imported into Surpac which has numerous internal validation routines, ensuring data integrity. Bedrock elevation and gravel thickness maps were created, primarily to identify first-pass targets and geological anomalies. The data were then used to create DTMs. Separate DTMs were created for the bottom and top contacts of the relevant gravel units. All boreholes that do not have gravel intersections were excluded from the model, as were boreholes with incomplete logs. The envelope around the boreholes were extended to one-half of the drill interval (to a maximum of 500 metres from a bulk-sample pit). Constraining strings (mined-out areas, farm boundaries, for example) were created. During this process all isolated outlier points were deleted from the data, in order not to introduce spurious errors. A block model was then created, based on the external limits of the DTM. The block size is based on the geology -- one by one by 0.3 metre was used for the Rooikoppie unit. No statistical manipulation or kriging is involved. All partial blocks were excluded from the model, resulting in a fairly conservative result. The model was then, further constrained by the two DTMs and the constraining strings, producing a volume. The figures for the mineral resources represent volumes available in the ground, fully depleted of material removed by the bulk-sampling programs.

Specific density

Specific gravity measurements are not routinely completed on alluvial gravels. SGs of gravels vary considerably from unit to unit due to moisture, clay and heavy mineral contents and conversion to tonnages tend to compound problems rather than simplify them. Regional averages for specific densities of alluvial gravels can vary from as low as 1.6 to over 2.4 where the gravels contain large percentages of banded iron formation. Typically, fluvial-alluvial gravels average 1.8 and Rooikoppie deposits are somewhat higher, at 2.0 to 2.2. All mining, processing and reconciliation on Rockwell's MOR projects is done using gravel volumes and not tonnages. However, for the purposes of individuals wishing to estimate tonnages, an average SG of 2.1 can be assumed for these deposits, as well as for other similar gravels along the MOR (the marginally higher value being due to the impact of the Rooikoppie gravels that are processed along with the fluvial-alluvial gravels).

Diamond grade

The grade of a diamond deposit is the estimated number of carats contained in a unit of weight or volume of gravel. Alluvial diamond grades are often measured in carats per unit of volume due to the problems associated with accurately predicting the bulk density of gravels, which are highly variable over short distances due to the relative pebble to sand content of the gravel and the pebble rock type.

In this news release, grades are reported in carats per hundred cubic metres and are based on grades recovered through Rockwell's production facilities. It is vital to note that different processing plants and/or recovery methods may not result in the same grades being recovered. In addition, it is noteworthy that, in this news release, mineral resource grades are based on bulk-sample, trial-mining and run-of mine recoveries -- there is no such thing as in situ grades, adjusted for processing parameters. Further, all recovered grades are based on a bottom cut-off of five millimetres.

Cut-off grades

Cut-off grades are not entirely pertinent to alluvial diamond mines since grade is not the only gauge of profitability. A far more relevant measure is the cut-off sieve/screen size, as diamond size more closely approximates value.

Further, fixed cut-off grades are not considered on Rockwell alluvial properties because diamonds do not have a fixed, single price, but their values vary significantly from one stone to another. Consequently, income per carat is extremely variable. As a result, the mine works on an income per unit of volume, rather than a simple cut-off grade. In addition, since Rockwell has a number of alluvial diamond mines in operation at any given time, individual operations may operate at a loss for a number of months before being put on care and maintenance or returning to profitability. Accordingly, companywide income per volume has to exceed cost per volume.

Cut-off screen sizes

One of the most fundamental issues on an alluvial diamond mine relates to the size distribution of both gravel clasts and diamonds. The details have important implications for top and bottom cut-off sizes on the screening plants and final recovery systems. Too-small screen-sizes at all locations will, ultimately, result in the loss of larger diamonds, but too-large screen-sizes will flood the plant with material. Although more stones may occur in the smaller size fractions, these do not make up the bulk of the value. Also, although the very large stones may be high value, they may make up so small a percentage of the population that the additional cost of processing larger size fractions may tender their recovery uneconomical. Consequently, any mining operation has an optimum range of stone size recovery where realized value exceeds input costs. The current bottom cut-off on the Remhoogte/Holsloot project is five millimetres.

The question of the size of diamonds passing through screens of certain apertures is extremely complex and relates to both the shape of the diamond as well as the smallest crosscut of the diamond which is presented at the screen. In addition, the behaviour of diamonds as they pass over a screen (in how many different ways they fall on the screen) is a further complicating issue.

Diamond value

Diamond values are quoted as U.S. dollars per carat (at a specified bottom cut-off, which must be the same as for the grade estimation). Valuation of the better-quality diamonds, which normally constitute most of the inherent value of the deposit, tends to be subjective and is best determined by sales values obtained on the open market in a free trading economy. The diamond values presented in this news release have been based upon the sale (not valuation) of over 10,000 carats during fiscal 2015.

Diamond size frequency distribution

The primary source of the MOR diamonds is projected to be the kimberlites in the catchments of the Vaal and Orange Rivers. This supposition is supported by size frequency distributions of alluvial deposits located on these rivers. The diamond population from the MOR, as typified by the over 35,000 carats recovered from the Brakfontein Hill complex of Rockwell's Saxendrift mine, can easily be derived from a combination of the Orange and Vaal river population. Large stones found along the MOR are likely to be derived from the Orange River and the smaller stones from both catchments.

The implication of these data is that the general SFDs of all of the MOR alluvial diamond deposits (downstream of Douglas) would be expected to be broadly similar, since no primary (kimberlite) sources feed diamonds into the MOR fluvial system downstream of the confluence of the Orange and Vaal Rivers at Douglas.

This assumption has been borne out by diamond recoveries by Rockwell on various terraces mined/prospected along the MOR. Minor differences in the SFDs of the various operations are to be expected, however, mostly due to variations in depositional environments.

As a consequence, the SFD of diamonds on the Remhoogte/Holsloot project is not expected to be significantly dissimilar from the other MOR properties. Minor variations will occur due to differences in local depositional environments as well as the specific percentage of Orange River versus Vaal River influence.

Steyn did not record the weights of each individual stone (but simply as daily parcels). Therefore, it has not been possible to confirm these details at this stage -- these data will be acquired by Rockwell during the forthcoming bulk-sampling program and presented in a forthcoming technical report.

Occurrence of large stones

Based on the SFD characteristics of any alluvial diamond deposit, it is possible to estimate the minimum volume of gravel that needs to be processed before diamonds of specific sizes should be recovered. An early study of processed volumes and large stone occurrences in some MOR operations showed reasonable correspondence at the company's Wouterspan and Saxendrift operations.

The general SFD of the MOR deposits was used to identify the expected stone count at different volumes processed. For MOR deposits, the first over-100-carat stone is only expected to be recovered after the processing of about 1.5 million cubic metres.

This press release was reviewed and approved and the data disclosed herein were verified by Rockwell's group technical manager, G.A. Norton, BScHons (PrSciNat), a qualified person who is not independent of the company, and Dr. T.R. Marshall, PhD (PrSciNat), a qualified person who is independent of the company.

We seek Safe Harbor.