The phenotype-linked fertility hypothesis suggests that females can judge male fertility by inspecting male phenotypic traits. This is because male sexually selected traits might correlate with sperm quality if both are sensitive to factors that influence male condition. A recent meta-analysis found little support for this hypothesis, suggesting little or no shared condition dependence. However, we recently reported that in captive zebra finches (Taeniopygia guttata) inbreeding had detrimental effects both on phenotypic traits and on measures of sperm quality, implying that variation in inbreeding could induce positive covariance between indicator traits and sperm quality. Therefore, we here assess empirically the average strength of correlations between phenotypic traits (courtship rate, beak colour, tarsus length) and measures of sperm quality (proportion of functional sperm, sperm velocity, sperm length) in populations of only outbred individuals and in mixed populations consisting of inbreds (F = 0.25) and outbreds (F = 0). As expected, phenotype sperm-trait correlations were stronger when the population contained a mix of inbred and outbred individuals. We also found unexpected heterogeneity between our two study populations, with correlations being considerably stronger in a domesticated population than in a recently wild-derived population. Correlations ranged from essentially zero among outbred-only wild-derived birds (mean Fisher's Zr ± SE = 0.03 ± 0.10) to moderately strong among domesticated birds of mixed inbreeding status (Zr ± SE = 0.38 ± 0.08). Our results suggest that, under some conditions, the phenotype-linked fertility hypothesis might apply.