Roughage provision as enrichment to piglets may contribute to reduce weaning associated problems such as fights at mixing with new conspecifics, diarrhoea and loss of body weight. The objective of this study was to investigate the effects of roughage provided at different phases to weaner piglets on body lesions (indicative of fights), diarrhoea and performance. Piglets (5 batches with each 48 piglets kept in 8 pens with each 6 piglets, N= 240 piglets) were weaned at around 24 days of age and kept in pens equipped with standard weaner diet, water and litter. Piglets varied regarding their experience with roughage, i.e., maize silage, resulting in 4 treatment groups: piglets received roughage before and after weaning (RR), only before weaning (RC), only after weaning (CR), or never (CC). Body lesions per piglet were counted before (reference) and after weaning, at day (d)1, d3 and d14 post-weaning (pw). Diarrhoea was evaluated individually and daily for 2 weeks pw by visually checking the presence of wet faeces in the anal area. Piglets were weighted the day of weaning, d3 and d14 pw and the average daily gain (ADG) was calculated between these periods. Body weight, ADG, number of body lesions and diarrhoea (yes/no) were analysed in R with generalized linear-mixed effects models with treatment, day pw and all possible 2-way interactions as fixed factors and piglet nested in maternity pen nested in batch and pw pen as random factors. Body weight at d14 pw differed between treatments (p< 0.05). Piglets in the CC and RC groups were heavier than piglets in the CR and RR groups (CC: 9.2 ± 0.2 kg; RC: 9.4 ± 0.2 kg, CR: 8.6 ± 0.2 kg and RR: 8.7 ± 0.2 kg). ADG was influenced by an interaction of treatment and day (p< 0.05). Between d3 pw and day14 pw, ADG was slightly higher for the CC (185.4 ± 34.8 g/day) and RC groups (187.5 ± 35.0 g/day) than for the CR (147.7 ± 34.8 g/day) and RR groups (113.1 ± 35.1 g/day). Number of body lesions per piglet was not affected by treatment (p>0.05). Diarrhoea was affected by treatment (p<0.05) as CR piglets had a 2.2 higher odds ratio for no diarrhoea then the average odds and the other groups. Piglets having access to roughage after weaning were lighter which could be due to a preference for feeding roughage rather than concentrate. Effects on diarrhoea seem only present in the CR group. Roughage did not affect the number of body lesions and thus did not seem to reduce fights at weaning. Planned behavioural analyses will provide more insight into the effect of roughage on weaner piglets behaviour, health and welfare at weaning.