Experimental evolution can be a useful tool for testing the impact of environmental factors on adaptive changes in populations, and this approach is being increasingly used to understand the potential for evolutionary responses in populations under changing climates. However, selective factors will often be more complex in natural populations than in laboratory environments and produce different patterns of adaptive differentiation. Here we test the ability of laboratory experimental evolution under different temperature cycles to reproduce well-known patterns of clinal variation in Drosophila melanogaster. Six fluctuating thermal regimes mimicking the natural temperature conditions along the east coast of Australia were initiated. Contrary to expectations, on the basis of field patterns there was no evidence for adaptation to thermal regimes as reflected by changes in cold and heat resistance after 1-3 years of laboratory natural selection. While laboratory evolution led to changes in starvation resistance, development time, and body size, patterns were not consistent with those seen in natural populations. These findings highlight the complexity of factors affecting trait evolution in natural populations and indicate that caution is required when inferring likely evolutionary responses from the outcome of experimental evolution studies.