Our study suggests that childhood pneumonia is prevalent and viruses, especially RSV, are detected in the majority of urban poor children aged <2 years with pneumonia in Bangladesh. A total of 77% of all the episodes of pneumonia in this cohort had a respiratory virus detected which is similar to findings from other studies of young children where etiological agents of lower respiratory tract infections have been thoroughly investigated 
confirming that respiratory viral pathogens are often associated with pneumonia in children <2 years of age 
. Most of the published literature from low and high income countries describe incidence of respiratory viral pathogen associated pneumonia among hospitalized children, including those from Sub-Saharan Africa where pneumonia is one of the leading causes of childhood death 
. Moreover, studies describing population based estimates have not included all the relevant respiratory viruses 
. So the findings from our study provide more comprehensive data on virus etiologies and community based incidence of pneumonia in very young children.
Similar to findings from other studies, RSV was the most frequently identified virus in this cohort of children with pneumonia 
. RSV infection in infants is frequently symptomatic, resulting in pneumonia and bronchiolitis in 30–70% of the cases 
. Our rate of RSV associated pneumonia of 12.5/100 child-year was comparable to incidence of RSV infection of 12/100 child-years and 15/100 child-years in children <1 years of age with pneumonia symptoms in rural Nigeria and Kenya 
. Furthermore the incidence of RSV associated pneumonia was highest among children aged less than 6 months. Similar to findings from our study, RSV has been associated with pneumonia and severe lower respiratory illness compared to other respiratory viral and bacterial pathogens in children from other low and middle income countries including Thailand, Indonesia, Gambia and Kenya 
. These lines of evidence suggest that for children <2 years RSV may be one of the major contributors of respiratory illness in both low and middle-income countries.
Influenza viruses and HMPV were also individually detected in 7% and 9% of episodes of pneumonia. Influenza virus infections were associated with 10% of childhood pneumonia among children aged <5 years in a study conducted in another urban setting of Bangladesh 
. Influenza virus infection in children may increase the risk of subsequent pneumococcal infection 
. HMPV is also associated with lower respiratory tract infection in previously healthy children 
. In a study conducted in Alaska, HMPV was more frequently identified in children <3 years of age hospitalized with respiratory infection than in children who did not develop respiratory symptoms suggesting that infection with the virus may be associated with severe illness 
. We also detected rhinoviruses in more than 10% of the pneumonia episodes. However, rhinoviruses can be detected in asymptomatic children and can be shed in nasal mucosa for several weeks after illness onset. Therefore we cannot confirm the association of the virus with illness in the cohort children 
In our study we observed increase in influenza activity in April–June of 2009 and 2010 which coincides with the influenza seasonality previously described in Bangladesh 
. However as the study was conducted only for two years and due to lack of nationally representative data on timing of increase activity of respiratory viruses other than influenza we could not establish seasonality for other viruses.
Our study has several potential limitations. Our study population was drawn from a single urban community therefore the data may not be generalizable to the whole country. Nevertheless our incidence of pneumonia in this cohort of children was comparable to other studies done in other low-income settings including urban Bangladesh 
. We only followed 515 children limiting our capacity to document the effect of respiratory viruses associated ARI and pneumonia on mortality. Our diagnosis of clinical pneumonia was not confirmed by chest radiographic examination and it is possible that children with bronchiolitis were diagnosed as pneumonia case-patients. In a resource poor setting it might not be feasible to confirm each episodes of clinical pneumonia by radiographic investigation and WHO definition of pneumonia is widely used in community setting to diagnose pneumonia. We also did not evaluate for bacterial etiologies. However the objective of the study was to identity the viral pathogens associated with respiratory disease in children and literature suggests that viral pathogen are predominant contributors of community acquired childhood pneumonia 
.We did not collect respiratory specimens from children without respiratory symptoms for comparison, limiting our confidence in implicating all of the identified viruses as causative agents of the respiratory illnesses. There is evidence that rhinoviruses and adenoviruses can be detected in asymptomatic children or shed for long periods of time after infection 
. Nevertheless, most of the viruses tested for, including RSV, HPIVs, HMPV and influenza viruses are seldom identified in asymptomatic controls 
The study findings demonstrated a high incidence of pneumonia associated with respiratory virus infection among children aged <2 years. These children were under close follow up and received timely treatment and referral services when necessary. Children under such close surveillance are likely to under represent severe illness 
. Other than influenza, there are no safe and effective vaccines to prevent childhood virus respiratory infections. Our data suggests that influenza infection was prevalent in children aged <6 months old, a group that could benefit from maternal immunization against influenza infection 
. Further research on developing safe and effective vaccines, especially for RSV, could also play an instrumental role in reducing the disease burden. In the meantime, in low-income settings research focused on developing cost-effective interventions to address modifiable risk factors such as improved hygienic behavior, air quality, breast feeding practices and nutrition 
may help reduce the overall burden of respiratory tract infection in children.