Scientists have created "autistic" mice by deleting a single gene in key parts of the brain.
The scientists were investigating what happens in brain cells
US researchers found the mice had traits such as poor social interaction and high sensitivity.
They say the findings, published in the Neuron, could point the way to better understanding of the causes of autism.
One UK expert said the findings were interesting, but needed much more work before they could be applied to humans.
An autistic spectrum disorder is a developmental disability that affects the way a person communicates and interacts with those around them.
Such disorders tend to emerge in childhood, and affect about 90 in 10,000 people. Boys tend to be affected more often than girls.
The University of Texas team looked at mice where the Pten gene - which has already been linked to other brain disorders - was deleted in the mature nerve cells in the cerebral cortex and hippocampus areas of the brain.
These regions are associated with higher brain function such as learning and memory.
The mice behaved in a number of socially abnormal ways, compared to another group of mice from the same litter.
The genetically altered mice were socially less skilled, being far less likely to be curious about new animals coming into the cage.
They also showed the same level of interest in an empty cage and in one containing another mouse - mirroring the behaviour of children with autistic spectrum disorders.
The genetically altered mice were also less likely to build nests or look after their young, but were more sensitive to stressful stimuli, such as loud noises or being picked up.
An examination of their brains showed they also had the increased brain volume and enlarged heads seen in people with autistic spectrum disorders
Dr Luis Parada, director of the Centre for Developmental Biology at the University of Texas Southwestern Medical Center in Dallas, who led the study, said: "It would be really exciting if it turned out that we've zeroed in on the anatomical regions where things go wrong in autistic patients, regardless of how the autism occurs."
Anthony Wynshaw-Boris and Joy Greer from the University of California San Diego School of Medicine in La Jolla, writing in the same journal, say the findings are "intriguing".
But they caution that the research does not provide the complete answer as there were other behaviours seen in people with autistic spectrum disorders - such as repetitive behaviours - which were not seen in the mice.
Professor Simon Baron Cohen, of the Autism Research Centre in Cambridge, added that the research " may have some relevance to understanding the genetic basis of autism spectrum conditions" because the mouse behaviour mirrored that of a subgroup of people with autistic spectrum conditions.
But he added: "Social abnormalities in a mouse may be caused by very different factors to human social abnormalities.
"Further human, clinical studies will be needed to test if PTEN is a susceptibility gene for autistic spectrum disorders.
"But this new study adds to our understanding of how genes expressed in the brain may have specific functions related to neuroanatomy and behaviour."