The relationship between genetic diversity and fitness has important implications in evolutionary and conservation biology. This relationship has been widely investigated at the individual level in studies of heterozygosity–fitness correlations (HFC). General effects caused by inbreeding and/or local effects at single loci have been used as explanations of HFC, but the debate about the causes of HFC in open, natural populations is still ongoing. Study designs that control for variation in the inbreeding level of the individuals, and knowledge on the function and location of the markers used to measure heterozygosity, are fundamental to understand the causes of HFC. Here we investigated correlations between individual heterozygosity and estimates of survival at different life-history stages in an open population of blue tits (Cyanistes caeruleus). For survival at the embryo, nestling and fledgling stage, we used a full-sibling approach, i.e. we controlled for the level of inbreeding. We genotyped 1496 individuals with 79 microsatellites mapped across 25 chromosomes in the zebra finch (Taeniopygia guttata) that were classified either as potentially functional (58 loci) or as neutral (21 loci). We found different effects of standardized multilocus heterozygosity (SH): SHfunctional had a negative effect on the probability of hatching and local recruitment of females, whereas SHneutral had a positive effect on adult survival. The negative effects of functional loci are better explained by local effects, whereas the positive effects of neutral markers could reflect inbreeding effects in the population. Our results highlight the importance of considering the characteristics of the markers used in HFC studies and confirm the mixed effects of heterozygosity in different contexts (e.g. sex and life-history stage).