Abstract Background Concerns about fussy eating are common amongst parents of young children. However, studies of the long-term impact of fussy eating show mixed results with regard to adequacy of dietary intake and child growth. This may be in part because there is no accepted definition of fussy eating and studies measure the construct in different ways, commonly relying on parent perception. This longitudinal analysis explores maternal and child characteristics associated with maternal perception of her toddler as a fussy eater in early toddlerhood and subsequent use of feeding practices at 2 years. Methods Mothers completed a self-administered questionnaire at child age 14 months, describing perception of their child as fussy/not fussy and child behaviour. Intake was assessed using a single 24-h recall and weight was measured by research staff. At child age 2 years mothers completed the validated 28-item Feeding Practices and Structure Questionnaire (FPSQ-28). Weight-for-age z-score (WAZ) was derived from WHO standards. Gram daily intake of fruit, vegetables and meat/alternative and a dietary diversity score were determined. Maternal/child characteristics independently associated (p ≤ 0.05) with perception of child as a fussy eater were determined using logistic regression. Variables were combined in a structural equation model assessing the longitudinal relationship between child/maternal characteristics, perception of child as a fussy eater and eight FPSQ factors. Results Mothers’ (n = 330) perception of her child as a fussy eater at age 14 months, was associated with higher frequency of food refusal and lower WAZ (R 2 = 0.41) but not dietary intake. Maternal perception as fussy (age 14 months) was associated with four FPSQ factors at 2 years (n = 279) - Reward for Eating, Reward for Behaviour, Persuasive Feeding and Overt Restriction, x 2 /df = 1.42, TLI = 0.95, CFI = 0.95, RMSEA = 0.04(0.03–0.05), PCLOSE = 0.99. Conclusions Lower relative child weight and food refusal prompted mothers to perceive their child as fussy. These behaviours in healthy weight children most likely reflect self-regulation of energy intake and neophobia. This perception was prospectively associated with use of non-responsive feeding practices, which may increase obesity risk. Future interventions could directly address perceptions of growth and fussiness, supporting parents to understand food refusal as developmentally appropriate behaviour in healthy young children. Trial registration ACTRN12608000056392 . Registered 29 January 2008.