Odonate assemblages were compared between replicate sets of shallow lakes that had been created and acidified by open-cast mining across a large area (2,451 ha) of southwest France (Arjuzanx, Landes); one set of lakes (n = 5) was experimentally restored by liming with calcium carbonate, whereas another group (n = 5) was left as untreated reference lakes. Both odonate adults and exuviae were sampled bimonthly during May–August 1998. Elevated turbidity and conductivity in limed lakes were the only physicochemical measures differing between restored and reference lakes, because deacidification occurred naturally, even in reference lakes during the 17 years after the onset of restoration. Restoration by liming can apparently lead to effects on lake turbidity that might be considered adverse. Twenty-four and 19 odonate species occurred among adults and exuviae, respectively, but there were no significant differences in richness between restored and reference sites. However, significantly, more exuviae were collected from the reference sites (588 vs. 180), where exuvial diversity and rank abundance indicated more evenly structured assemblages than those in restored lakes. Ordination showed that adult assemblages differed significantly between restored and reference lakes, and varied highly significantly with lake turbidity. This effect occurred because a small group of generally scarce adults were characteristic of reference sites (Chalcolestes viridis, Lestes virens, Cordulia aenae, Leucorrhinia albifrons, and Sympetrum sanguineum). Exuviae of these same species were less abundant at restored sites, but exuvial assemblages overall did not discriminate between restored and reference lakes. We conclude that lake restoration by liming can reduce diversity and larval numbers among odonates and subtly affects adult assemblages. In this case study, adult assemblages discriminated best between the lake types involved in the experiment, but important additional information arose from exuvial abundance and structure. This study indicates that natural recovery processes after acidification in formerly open-cast areas––rather than chemical intervention through liming––might lead to preferable conservation outcomes.