This paper presents a distributed random linear network coding approach for transmission and compression of information in general multisource multicast networks. Network nodes independently and randomly select linear mappings from inputs onto output links over some field. This approach achieves capacity with probability exponentially approaching 1 with the code length. It also performs compression when necessary in a network, generalizing error exponents for linear Slepian–Wolf coding. Benefits include decentralized operation and robustness to network changes or link failures. The approach can take advantage of redundant network capacity for improved success probability and robustness. It is illustrated that random linear network coding can offer advantages over routing in distributed network operation and networks with dynamically varying connections. The paper also provides a new bound on required field size for centralized network coding on general multicast networks. The paper discusses the model and preliminaries of network coding, including the basic model, algebraic network coding, and insights from bipartite matching and network flows. It presents success/error probability bounds for random linear network coding for independent and linearly correlated sources, and for arbitrarily correlated sources. It also gives examples of practical scenarios where randomized network coding can be advantageous compared to routing. The paper shows that random linear network coding achieves multicast capacity with probability exponentially approaching 1 with the length of code. It also demonstrates that random linear coding performs compression when necessary in a network, generalizing known error exponents for linear Slepian–Wolf coding. The approach not only recovers the capacity and achievable rates but also offers advantages such as decentralized operation and robustness to network changes or link failures. The paper also provides bounds on the probability of error-free transmission for independent or linearly correlated sources, which is tighter than the Schwartz–Zippel bound for general polynomials of the same total degree. The paper discusses the benefits of randomized coding over routing, including distributed settings and dynamically varying connections. It shows that random linear network coding can be particularly useful in these scenarios. The paper also presents simulations comparing distributed randomized coding to an approximate online Steiner tree routing approach, showing that random linear network coding outperforms the Steiner tree heuristic on a non-negligible set of randomly constructed graphs. The simulations also show that random linear network coding has lower overhead compared to routing.This paper presents a distributed random linear network coding approach for transmission and compression of information in general multisource multicast networks. Network nodes independently and randomly select linear mappings from inputs onto output links over some field. This approach achieves capacity with probability exponentially approaching 1 with the code length. It also performs compression when necessary in a network, generalizing error exponents for linear Slepian–Wolf coding. Benefits include decentralized operation and robustness to network changes or link failures. The approach can take advantage of redundant network capacity for improved success probability and robustness. It is illustrated that random linear network coding can offer advantages over routing in distributed network operation and networks with dynamically varying connections. The paper also provides a new bound on required field size for centralized network coding on general multicast networks. The paper discusses the model and preliminaries of network coding, including the basic model, algebraic network coding, and insights from bipartite matching and network flows. It presents success/error probability bounds for random linear network coding for independent and linearly correlated sources, and for arbitrarily correlated sources. It also gives examples of practical scenarios where randomized network coding can be advantageous compared to routing. The paper shows that random linear network coding achieves multicast capacity with probability exponentially approaching 1 with the length of code. It also demonstrates that random linear coding performs compression when necessary in a network, generalizing known error exponents for linear Slepian–Wolf coding. The approach not only recovers the capacity and achievable rates but also offers advantages such as decentralized operation and robustness to network changes or link failures. The paper also provides bounds on the probability of error-free transmission for independent or linearly correlated sources, which is tighter than the Schwartz–Zippel bound for general polynomials of the same total degree. The paper discusses the benefits of randomized coding over routing, including distributed settings and dynamically varying connections. It shows that random linear network coding can be particularly useful in these scenarios. The paper also presents simulations comparing distributed randomized coding to an approximate online Steiner tree routing approach, showing that random linear network coding outperforms the Steiner tree heuristic on a non-negligible set of randomly constructed graphs. The simulations also show that random linear network coding has lower overhead compared to routing.