Sex and environment may dramatically affect genetic studies, and thus should be carefully considered. Beginning with two inbred mouse strains with contrasting phenotype in the neuroma model of neuropathic pain (autotomy), we established a backcross population on which we conducted a genome-wide scan. The backcross population was partially maintained in small social groups and partially in isolation. The genome scan detected one previously reported quantitative trait locus (QTL) on chromosome 15 (pain1), but no additional QTLs were found. Interestingly, group caging introduced phenotypic noise large enough to completely mask the genetic effect of the chromosome 15 QTL. The reason appears to be that group-caging animals from the low-autotomy strain together with animals from the high-autotomy strain dramatically increases autotomy in the otherwise low-autotomy mice (males or females). The converse, suppression of pain behaviour in the high-autotomy strain when caged with the low-autotomy strain was also observed, but only in females. Even in isolated mice, the genetic effect of the chromosome 15 QTL was significant only in females. To determine why, we evaluated autotomy levels of females in 12 different inbred stains of mice and compared them to previously reported levels for males. Strikingly larger environmental variation was observed in males than in females for this pain phenotype. The high baseline variance in males can explain the difficulty in detecting the genetic effect, which was readily seen in females. Our study emphasizes the importance of sex and environment in the genetic analysis of pain.