Comparison of the SMA and OLS slopes showed that OLS regression significantly underestimated the slope between log body mass and log pronotum length in G. texensis (b SMA = 2.642, b OLS = 2.20; difference between slopes: t  =   2.92, df = 185, P  =   0.0039) and in A. domesticus (b SMA = 2.549, b OLS = 1.857; t  =   3.03, df = 116, P  =   0.003; Fig.1) suggesting that SMA better describes the scaling relationship between body mass and body size than OLS regression. Lower slope estimates (i.e., those derived from our OLS regressions) will tend to overestimate the condition of larger individuals (i.e., larger positive residuals for larger individuals; Fig.1). The SMA slopes for our reference populations were similar to those found in other studies on crickets (Kelly and Tawes 2013; Kelly et al. 2014). Although these slopes were lower than the value of 3.0 that is, predicted under simple geometric scaling (Green 2001) they were closer to 3.0 than the OLS estimates, which suggests that SMA better describes the ‘true” scaling relationship between M i and L i. Deviation from 3.0 is common across animals and arises because growth is rarely isometric (Peig and Green 2009).