In a world where human-robot collaborations are becoming increasingly important, facilitating trust in robots is essential. In humans, the level of interpersonal trust is partially determined through the sense of smell. Van Nieuwenburg et al. (2019) built on this idea and found that hexanal increases trust in humans. Based on their findings, the present preregistered study aimed to investigate whether hexanal could also increase the level of trust during human-robot collaboration. It was hypothesized that unmasked and eugenol-masked hexanal would increase the level of trust in human-robot interaction, suggesting a subconscious effect of hexanal on trust. These hypotheses were tested using a double-blind within-subjects design. Two experiments (n = 44 and n = 46) with sufficient statistical power to find the effects from Van Nieuwenburg et al. (2019) were conducted, serving as a direct replication of each other. Trust was operationalized using a visual detection task, where the human participant collaborated with a social robot. The participant indicated whether they had seen a target, after which they had the option to change their answer to the robot's. Trust was operationalized as the number of times participants changed their answer to the robot’s. Subjective trust was also measured using the Reliance Intention Scale. These tasks were performed over four odor conditions: hexanal, eugenol-masked hexanal, eugenol, and a neutral control condition. Contrasting previous work, a comparison of the conditions showed no significant effect of unmasked or eugenol-masked hexanal on the level of trust in robots. Effects of the odors on mood and arousal were ruled out. The findings are considered in the context of a potential mismatch between the natural smell of hexanal and the mechanical context of the robot. This raises new questions: would the effect be found in a different setting, with a different smell, or a more human-like robot?