Despite deleterious effects on individuals, the t haplotype is a selfish genetic element present in many house mouse populations. By distorting the transmission ratio, +/t males transmit the t haplotype to up to 90% of their offspring. However, t/t individuals perish in utero. Theoretical models based on these properties predict a much higher t frequency than observed, leading to the t paradox. Here, we use empirical field data and theoretical approaches to investigate whether polyandry is a female counterstrategy against the negative fitness consequences of such distorters. We found a significant decrease of the t frequency over a period of 5.5 years that cannot be explained by the effect of transmission ratio distortion and recessive lethals, despite significantly higher life expectancy of +/t females compared to +/+ females. We quantified life-history data and homozygous and heterozygous fitness effects. Population subdivision and inbreeding were excluded as evolutionary forces influencing the t system. The possible influence of polyandry on the t system was then investigated by applying a stochastic model to this situation. Simulations show that polyandry can explain the observed t dynamics, making it a biologically plausible explanation for low t frequencies in natural populations in general.