40Ar-39Ar laser probe ages are reported for individual clinopyroxene inclusions in thirty-two eclogitic diamonds from four localities: Jwaneng and Orapa (both in Botswana), Udachnaya (Siberia), and Argyle (Australia). Ages obtained from cleaved diamonds from Jwaneng (244 Ma), Orapa (102 Ma), and Udachnaya (425 Ma) are similar to those of their host kimberlites. However, there is considerable variation in age between individual laser extractions obtained from the same inclusion, which we suspect is mainly due to partial retention of radiogenic argon that diffused to the inclusion/diamond interface while the diamond was in the mantle. Some of this argon was lost when the diamonds were cleaved as part of the experimental procedure. The presence of an ambient mantle 40Ar component trapped in inclusions may be an additional cause of scatter in ages; however, this cannot be validated with the present data. In support of the diffusive loss process, the results from an Udachnaya inclusion completely enclosed within host diamond, with the argon being released by laser drilling, yielded an age of 1149 ± 37 Ma, which is considerably older than the host kimberlite (about 340-400 Ma). Likewise, the scatter in 40Ar K ratios for eclogitic diamonds from Jwaneng and Orapa indicates that they too are probably xenocrysts. Therefore, a combination of sample preparation technique and laser drilling can be used to offer a useful extension of the original method, enabling both diamond formation (from buried inclusions) and host rock emplacement ages (from cleaved inclusions) to be determined. Argyle clinopyroxenes, similarly cleaved to the diamond surface, also show a considerable range of inclusion ages (1159-1540 Ma); only in this case, there is a relationship between age and K Ca ratio of the inclusions. Such a relationship is difficult to reconcile in terms of partial loss of 40Ar and, instead, indicates that there was a protracted period of diamond growth.