Aims: To estimate the incidence and clinical characteristics in hospital admissions due to dehydration or undernutrition and their laboratory evaluation and treatment outcome in exclusively breastfed infants. Methods: All hospital admissions during the first 3 months of life assessed by the Dutch Paediatric Surveillance Unit (DPSU) between mid 2003 and mid 2005. Results: Nationwide 158 cases reported, correspond to an incidence of 58/y/100 000 breastfed infants; it is lower for severe dehydration at risk for hypernatraemia; 20/y/100 000. Sixty-five per cent of cases were <2 weeks old, their median weight loss was 9.3% and median age at admission 5 days; Serum sodium value was measured in only 12% of all cases. Insufficient volume intake and inadequate growth were most frequently reported (61% and 41%). Lethargy, jaundice or clinical dehydration was scored in 11-25%, seizures or shock in 3%. A breast pump at home was used in only 31%. In the hospital breast pumps were available (82%) as lactation consultants (73%). For treatment 65% was offered formula, in 30% by nasogastric drip. Most admissions lasted up to 3 days, all recovered fully and 33% were breastfed exclusively at discharge. Conclusion: The incidence of severe dehydration in the Netherlands is relatively low. With extended use of breast pumps at home it could be lower. To prevent complications, we recommend applying a reference weight chart, a full clinical examination and more extensive screening of serum sodium and glucose. ©2009 Foundation Acta Pædiatrica/Acta Pædiatrica.